ABSTRACT The purpose of the proposed work is to significantly advance the ability to precisely modify the zebrafish genome and consequently revolutionize how gene function can be studied using zebrafish. The technologies developed here will dramatically expand the kinds of experiments that can be performed and the kinds of questions that can be asked with the zebrafish. We modify the genome by 'gene targeting': a double strand break (DSB) induced by an engineered nuclease at a targeted locus is used to stimulate homologous recombination / homology directed repair between the targeted locus and a dsDNA donor molecule that harbors a modified version of the endogenous locus. Our current methods yield targeted modification of the zebrafish genome with very high efficiency: up to 1 in 6 of the treated animals transmit a precisely modified locus to offspring. Alterations on the order of 50-100 base pairs occur with precision at the highest frequency, whereas larger modifications, such as the introduction of a 1-2 kbp stretch of exogenous sequence, are recovered at lower rates. Here we develop tools for gene targeting using our current methods and new approaches to improve the efficiency and fidelity of gene targeting in zebrafish. Our first goal is to develop and test a set of tools and reagents that make it easy for any investigator to generate many standard types of modified loci in zebrafish. The toolkit we create will enable investigators to routinely produce: i) 'peptide knock-in alleles' in which peptide-encoding sequences have been integrated in frame with the normal coding sequence so antigen-tagged versions of the endogenous proteins are expressed; ii) 'reporter knock-in/knock-out alleles' in which the locus expresses a reporter protein instead of its normal product; iii) cre and creERT2 knock-in alleles; iv) 'bicistronic/two-product alleles' from which the endogenous product as well as a reporter are expressed; v) 'tagged' alleles in which a targeted change is identifiable by a co-introduced, tightly linked reporter gene; and vi) 'floxed conditional alleles', in which essential gene sequences are flanked by loxP recombination sites. A second goal is to develop and test a new approach for analyzing maternally supplied gene functions. Many studies using mutant zebrafish embryos to study signaling pathways, chromatin remodeling, or the control of pluripotency are confounded by the presence of maternally supplied gene products. We will generate reagents to accomplish conditional ablation of a gene in the germ line, allowing production of eggs and embryos that lack a maternally supplied gene product. Our final goal is to improve the efficiency and specificity of gene targeting. Our experiments will be aimed at i) stimulating targeting events that have increased probability of entering the germ line; ii) improving the efficacy with which longer sequences, including entire genes, can be introduced into the genome; and iii) improving the frequency of recovering floxed alleles, where two lox sites, often separated by >1 kbp need to be coordinately introduced. 1